Cancer of the colon and rectum is expected to be the third leading cause of cancer death in the United States in 2005. Hepatic metastases as the sole or dominant site of life-limiting disease occur in approximately 20-50% of these patients. Resection of these liver metastases results in five year survival approaching 40%, suggesting that regional control of this disease can lead to improved overall survival. Unfortunately, only a small fraction of patients with isolated liver metastases are eligible for resection. We are currently exploring a novel and highly promising approach to treatment of unresectable liver metastases, isolated hepatic perfusion. It is our goal to develop a regional hepatic therapy that has a response rate approaching 100% with a long duration of response. We envision then combining this successful regional treatment with an effective anti-tumor immunotherapy to ultimately have a significant impact on overall survival. This goal led to a phase I study of isolated hepatic perfusion with oxaliplatin for colorectal metastases. In this phase I study, out of six evaluable patients we had an objective radiographic partial response rate of 50%, a complete response rate of 17%, leading to an overall objective radiographic response rate of 67%. While this is a remarkable response rate for a phase 1 trial we do not believe that oxaliplatin alone is our optimal regimen for IMP. Given the profound synergy between 5-FU and oxaliplatin, we propose to examine the combination of 5-FU and oxaliplatin delivered via an isolated hepatic perfusion circuit. The rationale for this combination is rooted on a large body of in vitro and in vivo data suggesting significant synergy and no overlapping toxicity between these agents as well as important early clinical experience with systemic oxaliplatin that suggested that this agent was most effective when combined with other cytotoxic agents. [unreadable] [unreadable]